I think that FindFit package with the method Nminimize yields global minimum if initial values of adjustable parameters are chosen properly with some constraints. I did use it to fit one data set to one formula but I wonder if one can hint to me how I can make a simultaneous fit to two different data sets and two different equations with some shared parameters.

  • 2
    $\begingroup$ Please see this. $\endgroup$ Commented Aug 30, 2018 at 7:00
  • $\begingroup$ I tried two different simple polynomial equations with some common parameters to fit simultaneously two different sets of home-made data using both FindFit and NonLinearModelFit packages. Output values of common parameters were same and fits were good. But this was not the case when I have chosen more complex equations and the outcome oarameters and fits were bad and inconsistent. This will be a useless if one wants to fit simultaneously equations of real and imaginary parts of dielectric constant to their highly different measured data. This is a challange for both physicits and others !!!!! $\endgroup$ Commented Aug 31, 2018 at 9:48
  • 1
    $\begingroup$ When attempting to fit different data sets with some common parameters one has to apply the appropriate weight during the fitting process. See Simultaneous fit to three datasets with measurement errors using different models but same parameters for an example. $\endgroup$ Commented Aug 31, 2018 at 13:01


Browse other questions tagged or ask your own question.